List of experiments:
1. Create Perceptron Network to perform OR function.
2. Developing a multi-layer perceptron network and perfoming back propagation to classify images using Keras.
3. Building a regression multi layer perceptron using Sequential API and Functional API
4. Hyperparameter tuning in neural networks.
5. Avoiding Overfitting through different regularization techniques for neural networks.
6. Using Convolutional Neural Networks (CNN) for deep computer vision applicaitons.
7. Performing Image Classification using ResNet.
8. Implementing ResNet-34 using Keras.
9. Time series prediction using Recurrent Neural Network.
10. Time series prediction using Long Short Term Memory (LSTM) network.
11. EEG based seizure prediction using Gated Recurrent Unit (GRU) network.
12. Using WaveNet for handling long time series sequences.

To understand basic concepts related to deep neural networks.

To develop skills for using deep neural networks for healthcare applications.

To gain understanding of state-of-the-art tools and new architectures currently being used in deep neural networks.

After the completion of this course, the student will be able to:
1. Understand basics concepts related to deep learning.
2. Understand the different types of deep neural network architectures.
3. Develop deep learning models for classification and regression tasks related to healthcare applications.
4. Develop deep learning models for time series prediction, risk estimation for healthcare applications.
5. Optimise neural network models for improving accuracy and performance.

Ian Goodfellow, Yoshua Bengio and Aaron Courville, Deep Learning, The MIT Press, Cambridge Massacheusetts

Francois Chollet, Deep Learning with Python, Manning

Aurelien Geron, Hands-On Machine Learning with Sci-kit Learn, Keras and Tensorflow, O'Reilly